> For the complete documentation index, see [llms.txt](https://fennaf.gitbook.io/dataprocessing/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://fennaf.gitbook.io/dataprocessing/load-data-1/json.md).

# json

JavaScript Object Notation

The JSON format was inspired by the object and array format used in the JavaScript language. But since Python was invented before JavaScript, Python’s syntax for dictionaries and lists influenced the syntax of JSON. So the format of JSON is nearly identical to a combination of Python lists and dictionaries. Here is a JSON encoding that is roughly equivalent to the simple XML format: 

```
{"menu": {
  "id": "file",
  "value": "File",
  "popup": {
    "menuitem": [
      {"value": "New", "onclick": "CreateNewDoc()"},
      {"value": "Open", "onclick": "OpenDoc()"},
      {"value": "Close", "onclick": "CloseDoc()"}
    ]
  }
}}
```

Source: Python <https://docs.python.org/3.6/library/json.html>

{% hint style="info" %}
<https://www.youtube.com/watch?v=EfEm0g-bMPc>
{% endhint %}

a JSON object looks like a dictionary. It can have a dictionary as a value, and this can lead to a dictionary in a dictionary tree. It is derived from Javascript in which an object is described as a dictionary.&#x20;

### Load JSON

We can load JSON data into a python JSON object with the method`json.load()`. This is especially handy if we want only certain keys of the JSON file. In the example above I want to make a DataFrame of the hits with a record for each ID.  I load the entire JSON file and subtract the hits tree with the method `pd.DataFrame.from_dict()`

```python
import json
f = open('sample.json')
data = json.load(f)
data
```

```javascript
{'max_score': 5.9047804,
 'took': 47,
 'total': 288,
 'hits': [{'_id': '8660',
   '_score': 5.9047804,
   'entrezgene': '8660',
   'name': 'insulin receptor substrate 2',
   'symbol': 'IRS2',
   'taxid': 9606},
  {'_id': '3667',
   '_score': 5.812647,
   'entrezgene': '3667',
   'name': 'insulin receptor substrate 1',
   'symbol': 'IRS1',
   'taxid': 9606},
  {'_id': '3651',
   '_score': 5.288981,
   'entrezgene': '3651',
   'name': 'pancreatic and duodenal homeobox 1',
   'symbol': 'PDX1',
   'taxid': 9606}]}
```

I can investigate the structure of the JSON file by retrieving the keys. In the example below the key's `max_score`, `took`, `total` and `hits` are returned

```python
print(data.keys())
```

```python
dict_keys(['max_score', 'took', 'total', 'hits'])
```

It seems that the hits key contain interesting records we want to investigate further. These I will parse in a pandas DataFrame

```python
import pandas as pddf_data = pd.DataFrame.from_dict(data['hits'])
```

```python
df_data
```

| ​ | \_id | \_score  | entrezgene | name                               | symbol | taxid |
| - | ---- | -------- | ---------- | ---------------------------------- | ------ | ----- |
| 0 | 8660 | 5.904780 | 8660       | insulin receptor substrate 2       | IRS2   | 9606  |
| 1 | 3667 | 5.812647 | 3667       | insulin receptor substrate 1       | IRS1   | 9606  |
| 9 | 3651 | 5.288981 | 3651       | pancreatic and duodenal homeobox 1 | PDX1   | 9606  |

### Write to JSON <a href="#write-to-json" id="write-to-json"></a>

We can also define a JSON object and write it to a JSON file. We use the method `json.dump()` to do such

​

```python
data = {"menu": {
  "id": "file",
  "value": "File",
  "popup": {
    "menuitem": [
      {"value": "New", "onclick": "CreateNewDoc()"},
      {"value": "Open", "onclick": "OpenDoc()"},
      {"value": "Close", "onclick": "CloseDoc()"}
    ]
  }
}}

with open('output.json', 'w') as f:
    json.dump(data, f)
```

we can use [https://jsonlint.com](https://jsonlint.com/) to validate the output

​<https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_json.html>​

Mind you if you want to write a DataFrame to a JSON object you need to think about the orientation

```python
data = df_data.to_json(orient='records')
with open('output.json', 'w') as f:    
    json.dump(data, f)
```

[<br>](https://fennaf.gitbook.io/bfvm19prog1/data-processing/data-loading-storage-and-file-formats/xml)